Aleph Inference Gateway

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-3 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/infer -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "alchemi-bgr-nim", "atoms": [], "opttol": 0.02, "opttol_pressure": 0.7, "cellopt": false, "info": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-3 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/infer -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "alchemi-bmd-nim", "atoms": "value", "config": "value", "info": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "ancient-greek-bert", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "ankh", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "astrosage", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "aurora", "surf_vars": "<see input map>", "atmos_vars": "<see input map>", "lat": "<see input map>", "lon": "<see input map>", "time": "<see input map>", "atmos_levels": "<see input map>"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "aya-expanse-8b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biobert", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biogpt", "prompt": "Once upon a time", "max_tokens": 100}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biolinkbert", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biomed-roberta", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biomedbert-110m", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biomedbert-large", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "biot5", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-5 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/mit/boltz2/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "boltz-2", "polymers": [], "ligands": [], "constraints": [], "recycling_steps": 3, "sampling_steps": 50, "diffusion_samples": 1, "step_scale": 1.638, "output_format": "mmcif"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "borzoi", "sequence": "value", "n_bins": 16}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "chem-t5", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "chemberta-125m", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "chemgpt-1.2b", "smiles": "value", "max_new_tokens": 100, "num_return_sequences": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/energy -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "chgnet", "structure": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "chronos-bolt"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/embed -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "clay", "pixels": "<see input map>", "waves": "<see input map>", "gsd": "<see input map>", "lat": "<see input map>", "lon": "<see input map>", "time": "<see input map>"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "clinical-longformer", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "clinicalbert-110m", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-5 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/colabfold/msa-search/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "colabfold-msa-search", "sequence": "value", "sequences": [], "e_value": 0.0001, "iterations": 1, "databases": ["all"], "output_alignment_formats": ["a3m", "fasta"], "pairing_strategy": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "crysta-llm", "formula": "value", "temperature": 1.0, "max_new_tokens": 1024, "num_samples": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "deepseek-v2-lite-16b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~5-10 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "deepseek-v4-flash", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/dock -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "diffdock", "protein_pdb": "value", "ligand_smiles": "value", "num_poses": 10, "inference_steps": 20}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-3 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/molecular-docking/diffdock/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "diffdock-nim", "protein": "value", "ligand": "value", "ligand_file_type": "mol2", "num_poses": 10, "steps": 18, "time_divisions": 20, "save_trajectory": false, "skip_gen_conformer": false, "is_staged": false}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/embed -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "dinov3-vitl16", "image": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "dnabert-2-117m", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "enformer", "sequence": "value", "organism": "human"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "esmc-300m", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/structure -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "esmfold", "sequence": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~5-10 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/arc/evo2/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "evo2-40b", "sequence": "value", "num_tokens": 8, "temperature": 0.7, "top_k": 1, "top_p": 0.9, "enable_sampled_probs": false}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "gemma-3-4b-it", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/embed -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "geneformer"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-5 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/nvidia/genmol/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "genmol", "smiles": "value", "num_molecules": 30, "temperature": 1.0, "noise": 1.0, "step_size": 1, "scoring": "QED", "unique": false}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "geogalactica", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "glm-4-32b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "ithaca", "text": "text", "contextualize": false}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/images/generations -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "kandinsky-3", "prompt": "Once upon a time", "negative_prompt": "", "n": 1, "size": "1024x1024", "num_inference_steps": 25, "guidance_scale": 3.0, "seed": 1, "quality": "value", "image": "value", "strength": 0.75}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "lag-llama", "context": [], "prediction_length": 24, "num_samples": 100}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/retrieve -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "leandojo", "goal": "value", "num_premises": 5}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/design -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "ligandmpnn", "pdb": "value", "num_sequences": 1, "temperature": 0.1, "model_type": "ligand_mpnn"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "mace-mh-1", "elements": [], "positions": [], "lattice": [], "head": "omat_pbe"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "mace-mp", "elements": [], "positions": [], "lattice": [], "variant": "medium"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/energy -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "mace-mp-0", "structure": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "matscibert"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "mattersim", "predict": "value", "relax": "value"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "medgemma-27b-it", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~5-10 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "mistral-small-4-119b-2603", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "moirai", "values": [], "horizon": 12}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "moirai-large", "context": [], "prediction_length": 24, "freq": "h", "num_samples": 100}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "moirai-moe-1-0-r-base", "time_series": "<see input map>", "prediction_length": "<see input map>"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-5 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/nvidia/molmim/generate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "molmim", "smi": "value", "num_molecules": 10, "algorithm": "CMA-ES", "property_name": "QED", "minimize": false, "min_similarity": 0.7, "particles": 20, "iterations": 10, "scaled_radius": 1.0}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "moment", "time_series": [], "prediction_length": 96}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/embeddings -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "multilingual-e5-small", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "oceangpt-30b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "openbiollm-70b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-5 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/openfold/openfold3/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "openfold-3", "request_id": "value", "inputs": [], "diffusion_samples": 1, "output_format": "cif"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/science/forecast -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "pangu-weather", "date": "<see input map>", "lead_hours": "<see input map>", "coords": "<see input map>", "demo": "<see input map>"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "phi-4-reasoning", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "progen2", "prompt": "Once upon a time", "max_tokens": 256, "temperature": 1.0}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/translate -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "prostt5", "input": "."}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/design -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "proteinmpnn"}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-3 minutes)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/biology/ipd/proteinmpnn/predict -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "proteinmpnn-nim", "input_pdb": "value", "num_seq_per_target": 1, "sampling_temp": [0.1], "ca_only": false, "use_soluble_model": false}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~1-2 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "protgpt2", "prompt": "Once upon a time", "max_tokens": 120, "temperature": 0.7, "num_sequences": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "qwen25-vl-3b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "qwen25-vl-72b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~2-4 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "qwen25-vl-7b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~5-10 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "qwen3-235b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "qwen3-32b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done

Any request wakes it (0→1). The first call returns 503 model_scaled_to_zero with Retry-After; retry until 200 (OpenWebUI does this automatically).

# retry on 503 until the model is up (~3-6 min)
while ! curl -s -o /dev/null -w "%{http_code}" https://inference.vulcan.alliancecan.ca/v1/chat/completions -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" -d '{"model": "qwen36-27b", "messages": [{"role": "user", "content": "hi"}], "max_tokens": 1}' | grep -q 200; do sleep 5; done